How to Find Businesses Already Spending on Ads

Sooner or later every agency owner builds the same filter. Stop messaging businesses that might one day have a marketing budget, and go straight at the ones visibly spending money right now. If they are already paying for traffic, they have answered the hardest question in your sales process before you ever open your mouth: do you spend money on getting customers, yes or no.

The instinct is right. The execution is where people lose a week. "Already running ads" sounds like one fact, and it is actually four different claims stacked on top of each other — that an ad exists, that it is live now, that meaningful money is behind it, and that the person spending it is reachable and unhappy enough to talk to you. Most of the free tools prove the first two and say nothing about the last two, which is how you end up with a list of 300 businesses that all technically qualify and a reply rate that looks like any other list.

This article covers where ad spend is genuinely visible from the outside, what each method actually proves, the two opposite ways to read an active ad, and how to use the signal at list scale without it becoming a research job that eats the week you were supposed to spend selling.

What ad spend is really a proxy for

Nobody wants a prospect because they advertise. They want a prospect who can pay, who already believes in paid acquisition, and who is currently dissatisfied. Ad presence is a proxy for the first two and completely silent on the third.

It is a good proxy, and worth being precise about why. Most trigger signals you can detect from outside a business tell you about a problem: few reviews, a slow site, no booking form. A live ad is different, because it tells you about a decision. Someone in that company has already approved recurring spend against an acquisition goal, found the internal money for it, and set up whatever billing and tracking that required. That decision is the expensive part of your sale. Compared with a business whose owner has never once paid for a customer, you are starting several conversations in.

That is also the limit. A budget that exists is not a budget available to you, and a business that believes in paid acquisition has usually acted on that belief already — which brings us to the part of this filter people skip.

Where the spend is actually visible, and what each method proves

There are four ways to see ad activity from outside a company without paying for anything, plus a paid tier that packages them. They differ enormously in what they license you to conclude.

Meta's public ad library

Meta publishes a free, public, no-login archive of ads running across its platforms. You can search it by the advertising page or by keyword, filter by country, and see the creative itself along with when the ad started running. For an agency this is the highest-value free check available, because it shows you not just that a business advertises but exactly what it is saying, which is the raw material for any message you plan to send.

What the library will not tell you about an ordinary commercial ad is money. Spend and reach figures are published for the narrower category of ads about social issues, elections and politics, not for the plumber running lead-gen creative. So the library answers "are they advertising, and what is the pitch" and stays silent on "is this a few-hundred-dollar experiment or a serious monthly commitment." You can infer something from how many distinct creatives are live and how long they have been running — one ad up for a week is a test, nine variants running for four months is a program — but that is an inference, and you should label it as one in your own notes.

Google's Ads Transparency Center

Google runs a comparable archive for search, display and YouTube ads, searchable by advertiser name or by website domain. It shows creatives and formats and roughly when they ran. Three constraints matter for prospecting. It is keyed to advertisers who have completed Google's advertiser verification, so an unverified advertiser will not appear. It has no spend, keyword or targeting data. And you search it by advertiser, not by category — which means it is a tool for checking a company you have already named, not a tool for discovering companies.

Searching the money keyword yourself

Type the query your prospect's customer would type, in the geography they serve, and look at what carries a sponsored label. This is the fastest way to learn whether a category is being bid on at all, and it is the only method that shows you the competitive context rather than one advertiser in isolation. It is also the least complete: ad auctions rotate, results are personalised and geo-dependent, and you are seeing a snapshot of who won an impression for you at that moment, not a roster of everyone bidding.

Reading the destination

The page the ad points to tells you more about seriousness than the ad does. A dedicated landing page with a single offer, a conversion pixel and a form is the artifact of someone who has done this before, and the tags themselves are visible in any page's public source: an advertising pixel or conversion tag on a site is evidence that somebody set up measurement, which is a stronger signal of intent than a single live creative. An ad dropping traffic onto a homepage with the phone number in the footer is the artifact of someone spending money badly — which, if you sell conversion work, is the best prospect on the list rather than the worst. Nothing here requires anything exotic; it is looking at a public page the way any competitor would.

Paid tools

A market of tools scrapes the above, enriches it with contact data, and exports lists. They genuinely add coverage and save time. They also inherit every blind spot listed above — no tool can tell you a spend figure the platform does not publish — and they add a monthly cost that has to earn itself against a filter you can partly run for free.

Read that table once more with the real question in mind, because it is not the question the tools answer. What you want is not a flag saying "this business advertises." It is a threshold: is this business spending enough that the waste in it is worth somebody fixing? A few hundred dollars a month is not a client for most offers, because at that level the owner genuinely does not care whether it is spent well. A few thousand is a different conversation entirely, and it is the same business category, the same ad library entry, the same sponsored label. Presence is detectable from outside. Level is not. Everything you can do about that gap is a proxy: how many distinct creatives are live at once, how long the oldest has been running, whether the traffic lands somewhere built on purpose, whether the site carries conversion tracking at all. Stack three of those and you have a defensible guess at the tier. Treat the guess as a sort order, never as a qualification, and never quote a number back to the prospect as though you knew it.

The two opposite readings of an active ad

Here is the thing the tool marketing never says. Every live ad supports two contradictory conclusions, and which one applies decides your entire message.

Reading one: proven budget. They spend money to acquire customers. They will understand what you sell, they have a number in their head for what a customer is worth, and they do not need educating on the concept of paying for growth. This is the reading everyone has in mind when they build the filter.

Reading two: occupied seat. Somebody is already running those campaigns. It might be the owner's nephew, it might be a freelancer, it might be an agency two years into a retainer with a contract and a relationship. The ad is evidence of a decision already made and of an incumbent who made it.

Both are true at once, and the mix decides your odds. If you sell paid media management, reading two is your central problem: the businesses with the healthiest ad accounts are the hardest to win, and your real targets are the ones spending visibly badly. We go deeper on that whole motion in how to get PPC clients, and the industry view lives on our PPC agency page. If you sell something adjacent — SEO, web design, reputation, automation, lead generation — reading two barely hurts you, because you are not asking them to fire anyone. You are asking them to fund something the ad budget's performance makes obviously necessary.

Write your opener for the reading that applies to your offer. The generic "I noticed you're running ads" message is written for neither, which is why it converts like neither.

When ad presence is the wrong filter

The filter fails in three specific situations, and it is cheaper to recognise them now than after a campaign.

Your offer competes with the ads. If your pitch is some version of "get found without paying for every click," then heavy spenders are not automatically your best prospects — they are a specific argument, and one that a business happy with its return will dismiss in a sentence. The businesses worth your message are the ones spending reluctantly, and reluctance is not visible in an ad library.

Your category does not advertise on the platforms you can see. Plenty of real budget moves through channels with no public archive at all: trade publications, direct mail, sponsorships, conferences, referral fees. A profitable commercial contractor doing seven figures through relationships will show up nowhere in any ad tool and is not a worse prospect for it, just an invisible one.

You need volume. Ad presence is a narrow filter by design. Cut a niche by category, then by geography, then by ad activity, and the pool can collapse to a few hundred businesses — fine for a campaign you will run once, a problem for one you need to run every month. That tension between filter quality and pool depth is the whole subject of choosing qualifying filters for cold outreach, and it is worth reading before you commit to a filter this strict.

And sometimes the absence is the signal. A business ranking nowhere organically, running no ads, in a category where three competitors bid on its category name every day, is funding those competitors by default. That is a legitimate prospect with a legitimate problem — just a different message, and one you can only send if your list was not built on ad presence in the first place.

Why this doesn't scale the way the tools imply

Every method above is fundamentally an inspection: one advertiser, or one search result, at a time. None of them enumerates an area cleanly. Transparency archives index advertisers, not local businesses. Keyword searches surface whoever won an auction in that second. Even the paid tools are stitching together the same partial views.

Which means "build me a list of every business in Tampa running ads" is the wrong instruction, and chasing it is how prospecting turns into research. Invert the order instead:

1. Build the pool with filters that enumerate reliably. Category and geography do this. So do the business-profile signals that exist for every business whether or not it advertises — hours, review volume, website presence, whether anyone has claimed the listing.
2. Sample before you commit. Take thirty businesses out of that pool at random and check each one for ad activity. If two of the thirty are advertising, this filter cannot carry a campaign in that market. If eighteen are, it can, and now you know roughly what share of the pool it will keep.
3. Use ad presence as a score, not a gate. Tag the advertisers and message them first, with the message written for a proven budget. Keep the rest in the pool with a different message. You get the prioritisation benefit without the pool collapse.
4. Spend the deep research on the top slice only. Reading creatives and landing pages properly takes real minutes per business. Do it for the twenty you would most like to land, not for four hundred.

This is the same discipline that separates a signal-based list from a purchased one: the signal is the asset, and it has to be verified on the day you use it rather than inherited from a database built last year. The reputation-management version of the same method — two cheap public signals combined into one qualified trigger — is laid out in reputation management trigger signals, and the arithmetic generalises.

Saying what you saw without sounding like surveillance

Assume the prospect has received "I saw your Facebook ad" before. Assume it landed as spam, because in the version they got it was followed by nothing specific — the observation was the entire message, and it could have been sent to anyone advertising anything.

A reference to an ad earns its place only when it sets up something the recipient could not have predicted. Three tests before you send:

• Could this sentence have been sent to a thousand businesses? "I saw you're running ads" could. "You're running the same creative for water heaters and for whole-house repipes" could not.
• Does the observation connect to the thing you sell? If you cannot draw a straight line from what you noticed to what you do, the observation is decoration and it is costing you the reply.
• Would a stranger find this normal to know? Public ads and public websites are fair game and every competitor looks at them. Enumerating someone's spend, or implying you know their results, reads as creepy and invites an argument you cannot win because you do not actually have their numbers.

The strongest version is usually not a compliment or a critique but a question the ad makes obvious — the thing their campaign is plainly not built to handle. That is a live conversation, not a cold pitch, and it is why the message research half of this filter is worth more than the list-building half.

Proven budget still isn't a reachable person

One last step people skip, and it quietly wastes more money than a bad filter. Qualifying a business on ad activity says nothing about whether you can reach the human who controls that budget. A business can be advertising aggressively and still have no direct line published anywhere, a main number answered by a receptionist, and a contact form that routes to an inbox nobody reads.

So run the reachability check as a separate pass, after the fit check, and expect it to remove a meaningful share of the list — the mechanics of that shrinkage, and how to measure it on your own data before you pay for anything, are in what to do when half your prospect list has no contact details. The pay-capacity question is worth checking in the same pass, because "spends on ads" and "can afford your retainer" are different thresholds; which niches can actually afford your agency works through that properly.

None of this is difficult. It is just several passes, done in the right order, repeated every month as the data ages — which is exactly the work most agency owners intend to do and then don't, because it competes with client delivery. When we run outreach for an agency, that sequence is the product: find the businesses matching the profile using live business data and profile signals, validate that every number is a real working cell, then run the outreach and put the ready-to-buy ones straight onto the calendar. The filter is only useful if somebody actually runs it on Tuesday.

The short version

Ad spend is the cleanest evidence of budget you can get from outside a company, and it is free to check on Meta's public library, in Google's transparency archive, and in the search results themselves. None of those sources will tell you how much is being spent, and none of them enumerates a market, so use ad presence to prioritise a pool you built another way rather than to define the pool. Read every live ad two ways — proven budget, occupied seat — and write for whichever one your offer actually faces. Then check that the person behind the budget is reachable, because a perfect filter over unreachable contacts is an expensive way to send nothing.

Frequently asked questions